We are requesting support for the development of a three dimensional program to plan radiation therapy which will differ in a number of ways from available planning programs. The program will model treatments by external x- and Gamma-ray beams, electron beams, and brachytherapy sources. The program will feature a three-dimensional model of the patient and will emulate treatments by any combination of radiation sources and permit any feasible beam direction, so that non-coplanar beam directions can be evaluated. We propose to: develop techniques to synthesize disparate kinds of diagnostic data so that all available information can be used to develop a model of the patient's tumor and anatomy; provide flexible ways of marking and displaying 3D anatomic relationships; provide dose calculations in 3D for all conventional radiations available to us; develop techniques for the comparison in 3D of rival treatment plans; develop techniques to estimate and display uncertainty in the 3D dose distribution arising from uncertainty in patient position and other causes; and develop techniques to assist in the verification of treatments. We propose to conduct a clinical trial on at least 50 patients in which we will assess the benefit of the capabilities we will develop. In particular, we will assess the nature and extent of the benefit to the patient which non-coplanar treatments might offer. The goal of the research is to develop a responsive, rapid and easily used three-dimensional program which could be used clinically in planning the radiation therapy of cancer patients. This program could contribute significantly to reductions in treatment associated morbidity by making it possible to spare normal tissues from irradiation, and, by making possible higher tumor doses, could result in improved tumor control.